Free-Air Performance Tests of a5-Metre-Diameter Darrieus Turbine

Issued by Sandia Laboratories, operated for the United States

Department of Energy by Sandia Corporation.

NOTICE

This report was prepared as an account of work sponsored by

the United States Government. Neither the United States nor

the United States Department of Energy, nor any of their

employees, nor any of their contractors, subcontractors, or

their employees, makes any warranty, express or implied, or

assumes any legal liability or responsibility for the accuracy,

completeness or usefulness of any information, apparatus,

product or process disclosed, or represents that its use wouldnot infringe privately owned rights.

Printed in the United States of America

Available fromNational Technical Information ServiceU. S. Department of Commerce5285 Port Royal RoadSpringfield, VA 22161

Price: Printed Copy $4.50; Microfiche $3.00

sAND77-1063Unlimited Release

Printed December 1977

FREE-AIR PERFORMANCE TEST’S OF A 5-METRE-DIAMETER DARRIEUS TURBINE

Robert E. SheldahlBennie F. Blackwell

Aerothermodynamics Division 1333Sandia Laboratories

Albuquerque, New Mexico 87115

ABSTRACT

A 5-metre - diameter vertical-axis wind turbine has been tested at theSandia Laboratories Wind Turbine Site. The results of these tests andsome of the problems associated with free-air testing of wind turbinesare presented. The performance data obtained follow the general trendof data obtained in extensive wind tunnel tests of a 2-metre-diameterturbine. However, the power coefficient data are slightly lower thananticipated. The reasons for this discrepancy are explored in the paper,along with comparisons between experimental data and a computerizedaerodynamic prediction model.

ACKNOWLEDGMENTS

The authors are grateful for the support provided by the personnel of the Advanced Energy

Projects Division 5715 and R. E. Akins, 5443.

CONTENTS

Nomenclature

Summary

Introduction

The 5-Metre Vertical-Axis Wind Turbine

Testing and Data Acquisition

Results and Discussion

Conclusions

References

Figure

1

2

3

4

5

6

7

8

9

10

11

12

FIGURES

The Sandia 5- Metre Vertical Axis Wind Turbine at the Wind Turbine Site

The Present 3-Section Blades on the 5- Metre Turbine

Schematic of the 5- Metre Turbine System

Sample Record of Torque and Wind Speed

The Location of Two Anemometers with Respect to the 5- Metre Turbine

Wind Speed Frequency Distribution for the Four Data Sets

Power Coefficient,1

, Performance Data of the 5- Metre Turbine at150a and 125 rpm wit Wind Tunnel Data from a 2- Metre Turbine forComparison

power Coefficient, Kp, Performance Data of the 5- Metre Turbine at 150aand 125 rpm with Wind Tunnel Data from a 2-Met re Turbine for Comparison

C Performance Data of the 5- Metre Turbine at 150a, 125, 162.5, and1!?Obrpm

K Performance Data of the 5- Metre Turbine at 150a, 162.5, and1;Ob rpm

A Comparison of the 150a rpm Cp Data Between the Equatorial andTop Anemometers

A Comparison of the 162.5 rpm C+ Data Between the Equatorial andTop Anemometers

Page

7

9

11

11

13

16

28

32

Page

12

14

15

15

17

20

21

22

23

24

25

26

CONTENTS (cent)

Figure E%!%

13 A Comparison of the 150a rpm Kp Data Between the Equatorial and TopAnemometers 27

14 The 150a rpm 5- Metre Turbine Data Compared with Theory 29

15 The Zero Wind Drag Coefficient Data of the 5- Metre Turbine Comparedwith the 2- Metre Turbine and NACA-0012 Airfoil Section Data 30

Nomenclature

As

c

Turbine swept area

Blade chord

Cdo

CP

J

KP

L

m

n

n.1

N

Q

Qf

Qi

R

Rec

T.J

v m.1

x

Zero wind drag coefficient

QicdPower coefficient,

+ p&: A.

vm.

Advance ratio, ~

QiuPower coefficient,

+ p= A~(Ru)3

Blade length

Number of data records in a data set

Total number of data points in the data set

Number of data points in the ith velocity bin

Number of blades

Turbine torque

Friction tare torque

Average torque in velocity

Turbine maximum radius

bin i

Chord Reynolds number, -Pm

Turbine torque input to the BINS program) (Q + Qf)

Average freestream velocity in bin i

R@Turbine tip-speed ratio, ~

m.1

Nomenclature (continued)

Ifm Freestream viscosity

P. Reference density

Pa Freestream density

Lo Turbine rotational speed

NcLu Sol.idit y, ~

s

SUMMARY

The Sandia 5-metre vertical-axis wind turbine has been tested in free air at the Sandia

Laboratories Wind Turbine Site. The turbine was operated at nearly constant rotational speeds

by an induction motor/ generator which can act as either a motor delivering power to the turbine

or as a generator delivering power from the turbine to the utility line. The three blades on the

turbine consist of three sections: a circular arc located near the turbine equator with an

NACA-0012 airfoil cross section and two straight segments that attach the circular arc to the

axis of rotation. The solidity of this system is O. 26.

The performance data were obtained with the aid of a minicomputer and a computer program

which utilized statistical methods. The unsteadiness of the winds necessitated the statistical

averaging of the data. It was found that the wind turbine performance data are influenced by

gusty wind conditions, anemometer location, anemometer response time, and by the inertial

effects of the turbine rotor. The “method of bins” computer technique (computer code called

BINS) used for averaging the data is, at the present time, the only method by which reasonable

performance information has been obtained.

Four data sets for three constant rotational speeds were obtained. The results showed the

turbine’s performance to be less than anticipated from previous wind tunnel results of a 2-metre

turbine and also lower than computer calculations using an aerodynamic prediction model. The

maximum power coefficient, Cp, for the turbine was found to be O. 273. A computer analysis

shows that the performance of the turbine is degraded by the straight sections of the blades which

are not of airfoil cross section. Based on this, the performance data presented here should not

be considered to be representative of a vertical-axis wind turbine with blades which are of air-

foil cross section from hub-to-hub. Future tests of this 5-metre turbine will be conducted with

such blades.

9-1o

FREE-AIR PERFORMANCE TESTS OF A 5- METRE-DIAMETER DARRIEUS TURBINE

Introduction

The renewed interest in wind energy has brought about a reconsideration of ideas that

evolved when wind turbines were more fashionable. One such idea is the vertical-axis wind

turbine, 1 which was patented in the United States in 1931 by G. J. M. Darrieus. Sandia

Laboratories fabricated a 5-metre-diameter Darrieua wind turbine in 1974. The original tur-

bine design allowed for a variable rotational speed mode of operation. Subsequent studies

identified the constant rotational speed/ synchronous power grid application as being very

promising for the Darrieus turbine. Since early 1976, the Sandia 5-metre turbine has been

operating in a synchronous grid mode.

A limited amount of performance data was obtained on the turbine in 19762 but data

acquisition proved to be difficult and time consuming. At that time, strip charts of the turbine

torque and corresponding wind velocity were used to record the performance information. Early

in 1977 a minicomputer was installed and programmed to take the turbine performance informa-

tion and compute the power coefficients. This report presents the performance data obtained

to date.

The 5-metre turbine shares the site with a recently erected 17-metre turbine and a

2-metre turbine. Also at the site is a 30-metre-high meteorological tower and an instrumentation

building which houses the turbine controls, instrumentation, and Hewlett- Packard minicomputer.

The 5-Metre Vertical-Axis Wind Turbine

The Sandia 5-metre turbine, a proof-of-concept machine fabricated in 1974, was designed

to be erected in the shortest possible time at reasonable cost. The three -bladed turbine is

shown in Figure 1. Each blade consists of three sections: a circular arc located near the

turbine equator and two straight pieces that attach the circular arc to the axis of rotation. This

straight line/ circular arc combination was designed to approximate the shape that a perfectly

flexible blade would assume under the action of centrifugal forces and has been given the name

troposkien3 (Greek for turning rope). The curved part of each blade is a NACA-0012 airfoil

section with a 19-cm chord while each straight section is simply steel rolled to a “streamlined”

shape with a chord of 10 cm. The solidity, U, is calculated to be O. 26 assuming the 19-cm

chord of the circular arc is continuous to the axis of rotation.

11

Figure 1. The Sandia 5- NIetre Vertical Axis Wind Turbine at

the Wind Turbine Site.

12

The construction of each blade in three sections was done for cost considerations. It was

thought at the time that the attachment knuckles and the straight sections would have little in-

fluence on’ the operation and performance of the turbine. In Figure 2, the straight sections and

attachment knuckles are clearly visible. The differences in chord length between the circular

arc blade and the straight sections can also be seen. It is now believed that this construction

technique is detrimental to the performance of the turbine.

The turbine is designed to operate at nearly a constant rotational speed by connecting the

turbine shaft, through a two-stage timing belt drive, to an induction motor/generator operating

at 3500 rpm. By changing pulleys, the turbine speed can be changed in discrete steps. Figure 3

shows a schematic of the 5-metre system showing the relationship of the induction machine,

speed increaser, Lebow;~ RPM and torque transducer, and the turbine shaft. Nominal rotational

speed of the turbine is determined by the synchronous speed of the induction machine and relative

diameters of the gears. The induction machine can act as either a motor, delivering power to

the turbine from the utility line, or a generator, delivering power to the utility line from the

turbine.

Testing and Data Acquisition

The atmospheric testing of the 5-metre turbine with the minicomputer began in February

1977. The testing of turbines in free-air offers problems not usually encountered in wind tunnel

testing. In particular, the atmospheric wind speed seldom remains constant for any appreciable

length of time. Consequently, it is diffictit to decide the appropriate wind velocity corresponding

to a given torque measurement. A representative strip chart record of the wind velocity and

turbine torque is shown in Figure 4. The zero wind velocity and torque are at the baseline of

the record. The unsteadiness of the wind velocity and torque is quite evident and shows some of

the problems of obtaining free-air data from a wind turbine.

Banas4 and Sullivanz have developed a computer program named BINS which uses the

“method of bins” to statistically average the wind speed/torque data. The wind speed and torque

are recorded at sample rates, chosen by the operator, generally from 1 to 10 data samples per

second. The data are then stored in velocity bin widths of O. 5 miles per hour; i. e., a data point

is taken and the wind velocity is determined which in turn locates the velocity bin. The data point

is counted and the value of the torque is added to the total torque in the bin. The data are stored

as a function of the velocity bins (120 bins for velocities from O to 60 mph). Each bin records the

number of data points and the total summed torque. Each data record, consisting of the 120

velocity bins, number of data points, and the summed torque for each bins also contains informa-

tion which is constant for each data record. These constants are the rotational speed, number

,::Lebow Associates, 1728 Maple Lawn Road, Troy, Michigan 48084.

13

Figure 2. The Present 3-Section Blades on the 5- Metre

Turbine.

14

TimingSpeed

-b—

,Turbine Shaft

II I

1

1

Lebow‘Torque

BeltIncreaser

11’r

1 I

I }

I~lnduction Machine

RPM andTransducer

T’ 240 Volt, ldJ AC Power

Figure 3. Schematic of the 5- Metre Turbine System.

— Turbine Torque. . ..- . . ..~ind Speed

10 sec

ti

t , , t

Figure 4. Sample Record of Torque and Wind Speed.

15

of blades, anemometer identification, wind shear correction factor, temperatu re, barometric

pressure, time of the day, and the turbine tare torque. The turbine tare torque is the torque

lost in the turbine due to bearing friction and belt losses.

The computer will accept wind-velocity data from three separate anemometers; thus during

a single test, three data records can be generated, all with the same turbine torque information,

but with velocities corresponding to each anemometer. Figure 5 is a photograph of the 5-metre

turbine showing two anemometers. One anemometer is directly above the turbine, approximately

2 metres above the blade top attachment point on the turbine axis. The other anemometer is

located at the turbine equator height, approximately two turbine diameters south of the turbine

axis. That dimension and location were initially chosen because the availability of a corner fence

post facilitated mounting the anemometer. The southern direction was chosen because the pre -

vailing winds are generally from the east or west, and the anemometer would not be in the wake

of the turbine. The meteorological tower has four anemometers on it and is located approximately

20 turbine diameters west of the 5-metre turbine. The operator has the option of taking wind-

velocity data from any of the six available anemometers up to a total of three.

During a test, the required constant information is input to the computer. With the turbine

operating, the computer is instructed to take data. If during the test the temperature or barometric

pressure changes, the test is terminated and the data record stored. The new information is

input to a new data record and testing is resumed. Data are taken when the winds are available,

so a test may be a few minutes long or extend past an hour. These tests are then performed on

a day-to-day basis with the end result being a large amount of data taken for a wide range of wind

conditions over many days.

Results and Discussion

The data records for a given rotational speed and anemometer can be combined and the

performance of the turbine can be computed by the minicomputer in the control building. The

summed torque in each velocity bin, i, of a data record is

ni

T.=~Tj1 j=l

(1)

w’here ni equals the total number of data points in the velocity bin, i. When data records are

added together, the value of the torque in each bin cannot be directly added because the density

of the air when data records are taken will vary from day to day. Thus the value of the torque

must be adjusted to account for this using a reference density, po, and the actual freestream

16

Figure 5. The Location of Two Anemometers with Respect

to the 5- Metre Turbine.

density, pm, for each data record. Thus for each velocity bin, i, a normalized summed torque

is determined,

P.~i=T. —

1 Pm(2)

These values of ~i for each data record are summed along with the number of data points, ni,

for each velocity bin and an average torque for each velocity bin in a data set is determined as

~ (Ti)k

~i - kml

~ (ni)k

k= 1

where m is the number of data records being summed in the complete data set. A power

coefficient, which is a standard measure of performance,

Q,uC=l

P +P~3AOm. s

1

is then calculated by

(3)

(4)

where V~, is the average velocity of the velocity bin. The velocity bin width on each data record

is O. 5 mph. However, the operator may call for the data to be output in 1/2-, 1-, or 2-mph

increments. This is done by merely combining adjacent bins and again the velocity is the average

velocity of the wider increment; i. e., for a 1-mph increment in velocity, bin 1 and 2 wo~d com-

bine to form a bin with a 1-mph width and the average velocity would be O. 5 mph.

A second power coefficient has been defined by Banas4 as

QiuK=

p + poAs (Ru )3(5)

where the wind velocity has been replaced by the blade equatorial velocity. This power coefficient

was chosen by Banas for three reasons. They are: (1) Kp shows that power reaches a maximum

at a particular value of the advance ratio (wind speed) when the turbine rotational speed is con-

stant; (2) Kp describes more clearly the power output characteristics of the wind turbine operating

in the synchronous mode; and (3) since the calculation of Cp involves a wind velocity cubed, large

18

errors in the calculation can occur due to errors in the wind speed measurement. The tip speed

ratio is defined as:

.=*co.1

(6)

A nondimensional freestream velocity, called the advance ratio, is the inverse of Equation 6.

(7)

Data will be presented as Cp vs X and Kp vs J.

Four data sets were obtained for the 5-metre turbine. The first data set is for the constant

rotational speed of 150 rpm and was obtained in late February and early March 1977. This was

followed by a data set for 125 rpm obtained during late March, a data set for 162.5 rpm obtained

in June and July, and an additional data set again for 150 rpm obtained in late July and August 1977.

The wind speed frequency distributions of the four data sets are presented in Figure 6. The total

number of data points in each data set is n. This figure shows the percentage of data points in

each velocity bin and indicates how many data points are used to average the torque in each bin.

Where the frequency is low, the data may be suspect because only a few data points are used to

obtain an average.

The power coefficient, C data for the first two data sets are presented in Figure 7.P’

There are two data sets for 150 rpm; the early data set will be referred to as 150a rprn and the

latter set as 150b rpm. The 150a-rpm data set has a maximum Cp of O.273 while the 125-rpm

data set Cp is 0.213. These data are compared with the data5 obtained for a 2-metre turbinemax

of similar solidity tested in a wind tunnel. The wind tunnel data has a maximum Cp of O. 35 at a

lower (2. 85 x 105) Reynolds number than the Reynolds number of 4.0 x 105 for the 150a-rpm data.

The maximum Cp should increase with increasing Reynolds number.5 Thus the performance of

the 5-metre turbine in free-air is not aa good as anticipated. The 125-rpm data are also con-

siderably lower than the wind tunnel data and even lower than would be anticipated by the 15%5 5

reduction in Reynolds number to 3.4 x 10 from 4.0 x 10 for the 150a-rpm data. This has

caused considerable concern.

19

IWind Speed Frequency Distribution for the Four Data Sets

0.10

150a rpm

0.05-n ❑ 73,379

Oocmw

‘a- %

w

0.10I w125 rpm

n = 54,399000.05-

0

0 00 O@”oOo 000 0

0 000

0.O& %~n , r -

0.10I 162.5 rpm

0.05- 000 OoO”omn = 144,783

00°0

000 0

00 wo%oo

0.00. ,

15(II rpm

n = 88,295

0.10-

0

0.05- 0 @oOoo oooo-

%O@@ 000

0 000.00-P I I 1 1 1

0 10 20WIND VE?:CITY (MPfi

50 60

I 1 , , 1 Io 5 10 15 20 25

WIND VELOCITY (m/s)

Figure 6. Wind Speed Frequency Distribution for the Four Data Sets.

20

0.5

0.4

0.3

0.2

0.1

0.0

-0.1

5-m Turbine, u ❑ 0.26, Free-Air Data 2-m Turbine, o = 0.25, Wind Tunne

A 150a rpm, Rec ‘4. OX 105 0600 rpm, Rec = 2.8 x 105

❑ 125 rpm, Rec = 3.4 x 105

#o

0 &o A’

13”%~;ngnn

J“a.. .

A“””

A

ElA

13

El

o

0

A

❑

(3

A

o 1 2 3 4 5 6 7 8 9

TIP SPEED RATlO - X

Figure 7. Power Coefficient, ~, Performance Data of the 5- Metre Turbineat 150a and 125 rpm with Wind Tunnel Data from a 2-Metre Turbinefor Comparison.

Figure 8 presents the power coefficient, Kp, as a function of the advance ratio for the first

two data sets and compares them with the wind tunnel data of the previous figure. This figure

shows that the turbine power does reach a maximum for constant rotational speed as a function of

velocity. This figure also shows the inherent self-regulation of the Darrieus turbine operating at

constant rotational speed. One point of interest is the free-air data for advance ratios greater

than O. 3. The power produced by the 5-metre turbine does not decrease as rapidly past peak

Kp as the 2-metre turbine tested in the wind tunnel. This indicates that for larger advance ratios

there are some differences between the wind tumel information and data obtained in free-air

testing. Fortunately, this difference is in the form of improved performance for the turbine in

free-air.

21

8

7

6

5

4

3

2

1

A 150a rpm

0 125 rpm

o 600 rpm

(3

o %

o

&

($.“. 13

El

Eln

o

-1

-20.0 0.1 0,2 0.3 0.4 0.5 0.6

ADVANCE RATlO - J

Figure 8. Power Coefficient, Kp. Performance Data of the 5-Metre Turbine at 150aand 125 rpm with Wind Tunnel Data from a 2-Metre Turbine for Comparison.

A compilation of power coefficients, Cp, of all four data sets is shown in Figure 9. This

shows a most disturbing result. The data sets for 162.5 and 150b rpm are both lower in C thanP

the first data set of 150a rpm. It was because the 162. 5-rpm data set was found to be lower than

the 150a-rpm set that the 150a-rpm condition was repeated. Examining Figure 6 again shows

that the wind speed frequency distribution for the first data set (150a rpm ) is much smoother than

the other three sets. In reflecting back by memory, without benefit of having taken hard copy

data, it is believed that the winds were more steady for that data set. This conclusion, however,

is purely subjective and only the smoother wind speed frequency distribution of the first data set

can offer any additional support to that conclusion.

22

0.5

0.4

L0.3

0.2

0.1

0.0

-0.1

~oo0

““” %0QfP.....

K0.. : “

“.. ElfQ❑ %

●.D” El. . .

“ElEl

EJ

D&l

5-m Turbine, u= O.26

o 150a rpm, Rec ‘4. OX 105

c1 125 rpm, Rec ‘3.4x 105

0 162.5 rpm, Rec=4.1xlo5

d 150b rpm, Rec = 3.8 x 105

00

A

El

o 1 2 3 4 5 6 7 8 9

TIP SPEED RATlO - X

Figure 9. Cp Performance Data of the 5-Metre Turbine at 150a, 125, 162.5, and

150b rpm.

The power coefficients, Kp, for three data sets are shown in Figure 10. The data for all

three sets are nearly identical for advance ratios up to O. 3 where there is a divergence in the

data. This corresponds to velocities in excess of 26 mph for 150 rpm and 28 mph for 162.5 rpm;

the wind speed frequency indicates that the number of data points beyond those speeds drops

drastically, thus the averages may be affected. Assuming the divergence is real, then the

particular winds of the 162.5- and 150b- rpm data sets were in some way different from the winds

of the first (150a rpm) data set. This is not to say that the wind turbine’s output is a function of

the “quality” of the wind, but rather the problem may be the inability to correctly determine the

true wind velocity experienced by the turbine.

23

8

7

6

5

4

3

2

1

0

-1

-2

I I I I I

0 150a rpm

a 162.5 rpm

O 150b rpm

o&

f

. . .. .

%

. .0.” .,.

0 %0...0(3

am’

o

8

C@

ti.

0.0 0.1 0.2 0.3 0.4 0.5 0.6

ADVANCE RATl O - J

Figure 10. Kp Performance Data of the 5- Metre Turbine at 150a, 162.5,and 150b rpm.

24

As was mentioned previously, the minicomputer was programmed to accept wind speed

data from as many as three anemometers simultaneously. This capability was used extensively.

h ~gure 5, two of the anemometers are shown; one is 2-metres above the turb~e and the second

is at the turbine equator height located two turbine diameters south of the turbine axis. The third

anemometer that was used is located on the meteorological tower approximately 20 turbine

diameters to the west. No satisfactory data correlation was obtained from the third anemometer

primarily because of the distance between the turbine and anemometer. All data presented thus

far has used the anemometer located on the fence at the turbine equator height. Figure 11 shows

a comparison of the data for the two anemometers. It is immediately apparent that the top

anemometer is being affected by the turbine. As the tip speed ratio increases, the turbine be-

gins to look more like a solid barrier to the air and more of the air simply accelerates aromd it

causing the top anemometer to indicate a higher wind velocity. This translates to reduced power

coefficients. The anemometer on the fence is located. sufficiently far away so the influence of

the turbine, assuming an east or west wind, is less than 170of the anemometer indicated wind

speed. Figure 12 presents the same information as Figure 11 except it is for the 162. 5-rpm data

set.

I I I I I I I 1 I

0.5

0.4

0.3

0.2

0.1

0.0

-0.1

A

o

150a rpm

Equatorial Anemometer

Top Anemometer

~AAA A

o 1 2 3 4 5 6 7 8 9

TIP SPEED RATIO - X

Figure 11. A Comparison of the 150a rpm (+ Data Betweenthe Equatorial and Top Anemometers.

25

0.5

0.4

0.3

0.2

001

0.0

-0.1

—

162.5 rpm

A Equatorial Anemometer

o ToP Anemometer

0

A

Ao

A

1 1 I I I I c) 1 I

o 1 2 3 4 5 6 7 8 9

TIP SPEED RATlO - X

Figure 12. A Comparison of the 162.5 rpm Cp Data Between the Equatorialand Top Anemometers.

When the power coefficient, Kp. is examined for the two anemometers, the large discrepancy

is not obvious as shown in Figure 13. This indicates that when the accuracy of the wind velocity

is in question, the power coefficient, Kp, for constant rotational speed may be the better indicator

to use in assessing a turbine’s performance. Also, when two or more turbines are compared,

where there is some doubt about the accuracy of the wind measurement, K would be the betterP

criteria for making the comparison.

The question of why the 5-metre performance data are lower than expected based upon wind

tunnel data and theoretical modeling is still unanswered. In Figure 14, the best 5-metre turbine

data (150a rpm) are compared to two computed power coefficient, Cp, curves. AS noted pre-

viously, the three-section blades of the 5-metre turbine incorporate straight sections which are

of rolled steel and not airfoil sections. In the analysis, these sections were assumed to have

aerod~amic section characteristics similar to flat plates which have substantially higher drag

coefficients than the NACA-0012 airfoil sections. Curve 1 shown in the figure is obtained from

the multiple streamtube mode16 with empirical additions based upon wind tunnel results5 using

NACA-0012 airfoil section data for the curved blades and flat plate section data for the straight

sections. The calculated Cp curve is similar to but differs from the 5-metre turbine data.

Curve 2 shows the anticipated performance of the 5-metre turbine if the blades were airfoils from

hub-to-hub.

26

8

7

6

5

4

3

2

1

0

-1

150a rpm

A Equatorial Anemometer

O Top Anemometer

A

%0A

%0..J’

&LxYa6:

-2 ‘I I I I 1

0.0 0.1 0.2 0.3 0.4 0.5 0.6

ADVANCE

Figure 13. A Comparison of the 150aand Top Anemometers.

RATl O - J

rpm Kp Data Between the Equatorial

27

When the wind turbine is operated (powered) when there is no wind, a value for the zero

wind drag coefficient, Cdo. can be calculated. Under the no wind condition, the blades are always

operating at zero degree angle-of-attack. Figure 15 shows a comparison of the calculated Cdo’s

for the 5-metre turbine, 2-metre turbine, and the zero angle-of-attack drag coefficients for the

NACA-0012 airfoil section. The 2-metre model with continuous (hub-to-hub) airfoil blades

approaches the value for the NACA-0012 section data, while the Cdo’s for the 5-metre turbine

are greater by almost a factor of two. This indicates that there are large differences in the

blades between the 5- and 2-metre turbines. It is believed that the difference is primarily

caused by the “high drag” straight sections.

Conclusions

The performance data for the 5-metre turbine obtained with the aid of a minicomputer and

the computer program BINS shows the performance to be lower than anticipated from previous

wind tunnel test results on a 2-metre turbine and also lower than theoretical computer calculations.

A computer analysis, Figure 14, shows that the turbine performance is degraded significantly by

the straight sections of the blades which are not of airfoil cross section. The zero wind drag

coefficients, Figure 15, also indicate that the blades of the 5-metre turbine have a much higher

drag coefficient than anticipated. Based on this, the performance data presented here should not

be considered to be representative of vertical-axis wind turbines with blades which are of air-

foil cross section from hub-to-hub. Future tests of this 5-metre turbine will be conducted with

such blades.

In addition to the degradation of the performance caused by the straight sections, the data

collection methods may also cause the power coefficients, Cp, to be suppressed below their actual

values. The BINS program is, at the present time, the only method by which reasonable per-

formance information has been obtained. There are problems with this technique which are not

related to the BINS program but to the anemometry during gusty wind conditions. The anemometers

are much smaller than the turbine and are very responsive. They respond to a small volume of

air moving past them while the turbine must respond to a much larger volume of air which may

not be moving at a uniform velocity. This can lead to errors in recorded wind veloci~ since the

anemometer may not be responding to the same wind velocity which the turbine is responding to.

The anemometer also responds faster to increasing wind speeds than to decreasing wind speeds

with the net result on the average being a higher indicated wind speed than the actual average.

It has been shown in Figures 11 and 12 that anemometer location can have a large effect on the

turbine performance data. The anemometer must be far enough from the turbine to be unaffected

by the wind speeding up to go around the turbine at higher tip speed ratios and yet close enough to

be exposed to the same wind velocity and time variations in the wind as the turbine. Under gusty

conditions, it may not be feasible to satisfy both requirements for anemometer placement.

28

@ Computer Model with High DragStraight Sections

@ Computer Model with Airfoil SectiolBlades from Hub-to-Hub

0 150a rpm Data, Rec = 40 x 105

0 1 2 3 4 5 6 7 8 9

TIP SPEED RATIO - X

Figure 14. The 150a rpm 5- Metre Turbine DataCompared with Theory.

29

0.018

0.016

“=0 0.014

I

1-

E 0.012uiIbav~ 0.010&nc1zz~ 0.038pjN

O.W6

o

%

O 5-m Turbine

O 2-m Turbine

\NACA-0012 AirfoilSdion Data

105 2 4 6 8 lo6 2

CHORD REYNOLDS NUMBER - Rec

Figure 15. The Zero Wind Drag Coefficient Data of the 5-Metre TurbineCompared with the 2- Metre Turbine and NACA-0012 AirfoilSection Data.

30

The turbine is held at “nearly” constant rotational speeds; however, slip in the induction

machine will allow approximately a t 1-rpm variation of the turbine rotational speed. Any rapid

change in “the wind velocity will cause a slight change in rotational velocity; thus the inertia of

the turbine will affect the torque transducer output. The statistical averaging of the data by the

BINS program is assumed to average out inertizd effects.

The BINS computer code and the anemometry methods are being reexamined. It is

anticipated that the data collecting techniques can be improved to give truer turbine performance

data. The 5-metre tu:-bine has demonstrated its present capability and has successfully per-

formed its role as a proof-of-concept machine. Future testing of the turbine will be performed

with blades which have an airfoil cross section from hub-to-hub. A 20~0or greater increase in

performance is expected with the new blades.

31

References

1. Blackwell, B. F., “The Vertical Axis Wind Turbine ‘ HOWit Works’, “ SLA-74-0160#Sandia Laboratories, Albuquerque, NM, April 1974.

2. Sullivan, W. N. , “5-M Turbine Field Testing, ‘‘ Proceedings of Vertical-Axis WindTurbine Technology Workshop, SAND7 6-5586, Sandia Laboratories> Albuquerque NMsMay 1976, Section IL PP 107-117.

3. Blackwell, B. F., and Reis. G. E., “Blade Shape for a Troposkien Type of Vertical-AxisWind Turbine, ” SLA-74-01 54, Sandia Laboratories, Albuquerque, NM* April 1974.

4. Reuter, R. C. , and Sheldahl, R. E. , (editors) “Sandia Laboratories Vertical-Axis WindTurbine Program Technical Quarterly Report, April-June 1976, SAND76-0581, SandiaLaboratories, Albuquerque, NM, November 1976, pp 24-31.

5. Blackwell, B. F., Sheldahl, R. E., and Feltz, L. V. , “wind Tunnel Performance Data forthe Darrieus Wind Turbine with NACA-0012 Blades, “ SAND76 -0130, Sandia Laboratories,Albuquerque, NM, May 1976.

6. Blackwell, B. F., et al, “Engineering Development Status of the Darrieus Wind Turbine. “Journal of Energ~, Vol. I, No. 1, Jan. -Feb. 1977, pp 50-64.

32

DISTRIBUTION :

Wichita State University (2)Aero Engineering DepartmentWichita, KS 67208Attn: M. Snyder

B. Wentz

D. K. Ai.Sr. Scientific Assoc.Alcoa LaboratoriesAluminum Company of AmericaAlcoa Center, PA 150~9

American Wind Energy Association21243 Grand RiverDetroit, MI 48219

Peter AndersenKarstykket 42Vvelse 3550Slagerup, Denmark

South Dakota Sch. of Mines & Tech.Dept. of Mechanical EngineeringRapid City, SD 57701Attn: E. E. Anderson

P. BaileyP. O. Box 3Kodiak, AK 99615

Washington State UniversityDepartment of Electrical EngineeringCollege of EngineeringPullman, WA 99163Attn: F. K. Bechtel

Arizona State UniversitySolar Energy CollectionUniversity LibraryTempe, AZ 85281Attn: M. E. Beecher

University of OklahomaAero Engineering DepartmentNorman, OK 73069Attn: K. Bergey

Louisiana Tech University (1O)Dept. of Mechanical EngineeringRuston, LA 71270Attn: B. F. Blackwell

Dr. P. H. BottelberghsChemical Conversion & Energy StorageLandelijke Stuurgroep

Energie OnderzoekDutch National Steering Group

for Energy ResearchLaan van Vollenhove 3225Zeist (NETHERLANDS)

McDonnell Douglas Aircraft Co.P. O. BOX 516St. LOUiS, MO 63166Attn: R. Brulle, Dept. 241, Bldg. 32

P. CalnsnElectrical Research AssociatesCleeve RoadLeathe rheadSurrey, England

R. CamereroFaculty of Applied ScienceUniversity of SherbrookeSherbrooke, QuebecCanada JIK 2R1

J: Carter1420 Felts RoadHealdsburg, CA 95448

V. A. L. ChasteauDepartment of Mech. EngineeringThe University of AucklandPrivate BagAuckland, New Zealand

E. N. ChesbroBOX 162Londonderry, VT 05148

University of HawaiiWind Engineering Research DigestSpalding Hall 357Honolulu, HI 96822Attn: A. N. L. Chiu

USCG R & D CenterAvery PointGroton, CT 06340Attn: Lt. T. Colburn

Lockheed California Co.BOX 551-63A1Burbank, CA 91520Attn: U. A. Coty

US Department of Ehergy (2)Nevada Operations OfficeP. O. Box 14100Las Vegas, NV 89114Attn: R. Ray, Operations

H. Mueller, ARL

33

DISTRIBUTION:( (cent)

US Department of Energy (3)Albuquerque Operations OfficeP. O. Box 5400Albuquerque, NM 87115Attn: D. K. Knowlin

W. P. GraceD. W. King

Department of Energy Headquarters (2O)Wind Energy Conversion Br.Washington, DC 20545Attn: L. Divone, Chief

D. D. Teague

Southern Illinois UniversitySchool of EngineeringCarbondale, IL 62901Attn: C. W. Dodd, Elec. Sciences and Sys. Eng.

Electric Power Research Imstitute3412 Hillview AvenuePalo Alto, CA 94304Attn: P. Box, Program Manager, Solar Energy

J. FischerF. L. Smidth & Company A/SVigerslevalle 772500 Valby, Denmark

Robert E. FisherEnvironmental Protection SpecialistDepartment of Environmental Resources736 West Fourth StreetWilliamsport, PA 17701

Aerophysics Company3500 Connecticut Avenue NWWashington, DC 20008Attn: W. F. Foshag

Albert Fritzs cheDornier System GmbHPostfach 1360’7990 FriedrichshafenWest Germany

Amarillo CollegeAmarillo, TX 79100Attn: E. Gilmore

TRW Energy Systems7600 Colshire DriveMclean, VA 22101Attn: R. German

Solar Central7213 Ridge Rd.Mechanicsburg, OH 43044Attn: D. Greider

Massachusetts Institute of Technology77 Massachusetts AvenueCambridge, MA 02139Attn: N. D. Ham

Massachusetts Inst. of TechnologyAero/Astro DepartmentCambridge, MA 02139Attn: W. L. Harris

Irwin Industries6045 West 55th PlaceArvada. CO 80002Attn: G. P. Hawkins

Jet Propulsion Lab4800 Oak Grove DrivePasaderia, CA 91103Attn: G. Herrera

Allegany Ballistics LaboratoryHercules IncorporatedP. o. Box 210Cumberland, MD 21502Attn: P. W. Hill

Sven HugossonBOX 21048S. 100 31 Stockholm 21SWEDEN

O. IgraDept. of Mechanical EngineeringBen-Gurion University of the NegevBeer- Sheva, Israel

J. B. F. Scientific Corporation2 Jewel DriveWilmington, MA 01887Attn: E. E. Johanson

Stanford UniversityMechanical EngineeringStanford, CA 94305Attn: J. P. Johnston

Kaman Aerospace CorporationOld Windsor RoadBloomfield, CT 06002Attn: J. Barzda

Chief of Systems Research

Michigan State UniversityDivision of Engineering ResearchEast Lansing, MI 48824Attn: O. Krauss

Lawrence Livermore Laboratory (2)P. O. )30x 808 L-340Livermore, CA 94550Attn: D. W. Dorn

D. Hardy

34

DISTRIBUTION: (cent)

Public Service Co. of NMP. O. BOX 2267Albuquerque, NM 87103,4ttn: M. Lechner

State Energy CommissionIResearch and Development Division:1111 Howe Avenue;Sacramento, CA 95825Attn: J. Lerner

‘[TS Department of AgricultureAgriculture Research CenterBuilding 303Beltsville, MD 20705.Attn: L. Liljdahl

D. LindleyUniversity of CanterburyMechanical EngineeringChristchurch, New Zealand

Aeroenvironment, Inc.660 South Arroyo ParkwayPasadena, CA 91105Attn: P. B. S. Lissaman

One LjungstromSwedish Board for Tech. DevelopmentFACKS-1OO 72 Stockholm 43, Sweden

Los Alamos Scientific LabP. O. BOX 1663Los Alamos, NM 87544Attn: J. D. Balcomb Q-DO-T

Library

L. H. J. MaileEngineering ManagerDominion Aluminum Fabricating Ltd.3570 Hawkestone RoadMississauga, Ontario L5C 2V8CANADA

Daedalean Associates15110 Frederick RoadWoodbine, MD 21797Attn: F. Matanzo

Kaman Sciences Corp.P. O. BOX 7463Colorado Springs, CO 80933Attn: J. Meiggs

Dept. of Economic Planning andDevelopment

Barrett BuildingCheyenne, IVY 82002Attn: G. N. Monsson

Mineral Development Geologist

National Aeronautics and Space Adm.Langley Research CenterHampton, VA 23665Attn: R. Muraca, MS 317

National Aeronautics and Space Adm. (3)Lewis Research Center2100 Brookpark RoadCleveland, OH 44135Attn: J. SaVino, MS 509-201

R. L. ThomasW. Robbins

West Texas State UniversityDepartment of PhysicsP. O. BOX 248Canyon, TX 79016Attn: V. Nelson

J. Nightingale1735 Hunt AvenueRichland, WA 99352

Oklahoma State University (2)Stillwater, OK 76074Attn: W. L. Hughes, EE Dept.

D. K. McLaughlin, ME Dept.

Oregon State University (2)ME DepartmentCorvallis, OR 97331Attn: R. Wilson

R. W. Thresher

HelionP. O. Box 4301Sylmar, CA 91342Attn: J. Park

Northwestern UniversityDepartment of Civil EngineeringEvanston, IL 60201Attn: R. A. Parmelee

B. Maribo PedersenDept. of Fluid MechanicsBldg. 404, DTH2800 LyngbyDENMARK

Colorado State UniversityDept. of Civil EngineeringFort Collins, CO 80521Attn: R. N. Meroney

35

DISTRIBUTION: (cent)

A RobbMemorial Univ. “of NewfoundlandFaculty of ~g. & Applied SciencesSt. John’s NewfoundlandCandda AIC 5S7

Stanford Electronic LaboratoriesRadio Science LaboratoryStanford, CA 94305Attn: A. V. da Rosa

H. SevierRocket and Space DivisionBristol Aerospace Ltd.P. O. BOX 874Winnipeg, ManitobaR3C 2S4 Canada

P. N. ShankarAerodynamics DivisionNational Aeronautical LaboratoryBangalore 560017India

Cornell UniversitySibley School of Mechanical and

Aerospace EngineeringIthaca, NY 14853Attn: D. G. Shepherd

Learning Unlimited Corporation72 Park StreetNew Canaan, CT 06840Attn: K. K. Skeele

Iowa State UniversityAgricultural Engineering, Rm 213Ames, IA 50010Attn: L. H. Soderholm

Southwest Research Institute (2)P. O. Drawer 28501San Antonio, TX 78284Attn: W. L. Donaldson

Senior Vice PresidentR. K. Swanson

Iowa State UniversityClimatology and Met erology312 Curtiss HallAmes LA 50010Attn: E. S. Takle

R. J. Templin (3)Low Speed Aerodynamics SectionNRC-National Aeronautical EstablishmentOttawa 7, Ontario, Canada K1AOR6

Texas Tech University (3)P. O. BOX 4289Lubbock, TX 79409Attn: K. C. Mehta, CE Dept.

J. Strickland, ME Dept.J. Lawrence, ME Dept.

United Engineers and Constructors, Inc.Advanced Engineering Department30 South 17th StreetPhiladelphia, PA 19101Attn: A. J. Karalis

Manager. Special Products

United Nations Environment Programme485 Letington AvenueNew York, NY 10017Attn: I. H. Usamani

University of New Mexico (2)Albuquerque, NM 87106Attn: K. T. Feldman

Energy Research CenterV. Sloglund

ME Department

Illinois Institute of TechnologyDept. of Electrical Engineering3300 South FederalChicago, IL 60616Attn: A. G. Vacroux

Otto de VriesNational Aerospace LaboratoryAnthony Fokkerweg 2Amsterdam 1017The Netherlands

West Virginia I.. DiversityAero Engineering1062 Kountz AveMorgantown, WV 26505Attn: R. Walters

Bonneville Power AdministrationP. O. BOX 3621Portland, OR 97225Attn: E. J. Warchol

D. F. WarneERA LimitedCleeve RoadLeatherheadSurrey KT22 7SAENGLAND

36

DISTRIBUTION: (cent)

Tulane UniversityDept. of Mechanical EngineeringNew Orleans, LA 70018Attn: R. G. Watts

University of AlaskaGeophysical InstituteFairbanks, AK 99701Attn: T. Wentink, Jr.

Shell Oil CompanyEnergy Economics & Forecasting

planning & EconomicsP. O. BOX 2463Houston, TX 77001Attn: S. M. Lambert, Manager

Dynergy Corp.P. O. BOX 4281269 Union AvenueLaconia, NH 03246Attn: R. B. Allen, Gen. Manager

T. P. RomeroP. O. BOX 2806Las Vegas, NM 87701

DOE/ DSE20 Massachusetts Ave.Washington, DC 20545Attn: S. Hanson

W. W. Garth, IVTyler & Reynolds & CraigOne Boston PlaceBoston. MA

Edward ChesbroLondonderry, VT 05148

James D. Foch, Jr.Department of Aerospace

Engineering SciencesUniversity of ColoradoBoulder, CO 80309

Morey/ Stjernholm & Associates1050 Magnolia St.Colorado Springs, CO 80907Attn: D. T. Stjernholm, P. E.

Mechanical Design Engineer

100012001260130013201324132413301331

G. A. FowlerL. D. smithK. J. TouryanD. B. ShusterM. M. NewsomE. C. RightleyL. V. FeltzR. C. MaydewH. R. Vaughn

1332 C. W. Peterson1333 S. McAlees, Jr.1333 R. E. Sheldahl (20)1334 D. D. McBride1335 W. R. Barton1336 J. K. Cole3161 J. E. Mitchell (50)3161 P. S. Wilson5333 R E. Akins5333 J. W. Reed5700 J. H. Scott5710 G. E. Brandvold5715 R. H. Braasch5715 E. G. Kadlec (100)5715 B. Stiefeld5715 W. N. Sullivan8266 E. A. Aas3141 C. A. Pepmueller (Actg ) (5)3151 W. L. Garner (3)

For DOE/ TIC (Unlimited Release )DOE/TIC (25)

(R. P. Campbell. 3172-3)

37